De Novo Engineering of Small Molecule-Actuatable Biosensors for Cell Therapy ABSTRACT Preceded by small-molecule drugs and biologics, cell-based therapies, such as the use of chimeric antigen receptor T (CAR-T) cells against blood cancers, is becoming a new pillar of medicine. However, compared with traditional drugs, they are more susceptible to safety concerns due to difficulties associated with controlling cell actions in a therapeutic setting. My research program aims to develop a novel approach to engineer genetically encoded biosensors for spatiotemporal control of cell behaviors. We focus on the de novo engineering of the biosensors for using small molecules to control cell therapeutic responses, proliferation, death, migration, communication, and metabolism. Our strategy is to design chemically induced dimerization (CID) systems, in which two proteins dimerize only in the presence of a small molecule. To date, only a few CID systems are available and their dimerization inducers are not ideal for clinical use. The de novo design of CID systems with desired affinity and specificity for given small molecules remains an unsolved problem in the field of protein engineering. We will solve this problem by coupling computational protein design to our recently developed single-molecular-interaction sequencing (SMI-seq) technology. SMI-seq has the potential to break a key barrier to CID engineering by enabling large-scale, two-dimensional (or ?library-by-library?) screening of two CID binder variant libraries. We will apply two parallel approaches to engineer CID systems: i) targeted screening of computationally designed binder libraries, and ii) random screening of vastly diverse combinatorial binder libraries (>109) using immunoglobulin or de novo designed scaffolds. We will assess the success rates, turnaround times, and cost-effectiveness of both approaches by testing a set of clinically approved antiviral drugs with excellent intracellular delivery efficiency as CID inducers. Finally, we will demonstrate the use of designed biosensors to control cellular processes in both cultured cells and a mouse model. Successful completion of this research will open up new possibilities for engineering ligand-responsive protein assemblies, an unexplored territory of protein design. Designed CID systems will significantly expand the chemogenetic toolkit for gene- and cell-based therapies, as well as systems and synthetic biology research.